Monday, January 1, 2007

research topics

A couple of my current research topics:

1) Whether I should (and if the answer is yes, then how to) rehost Graphic-Forms on the .NET Framework, thus abandoning User32/GDI?

RDNZL is proof that Common Lisp and .NET integration is possible. Dr. Weitz certainly got people's attention with that project (which as he points out was heavily inspired by earlier work on an earlier Scheme-based project). I also discovered that the SWT folks have proposed doing a port of that Java-based GUI library. So feasability does not seem to be in doubt. The question is whether it makes sense for my project.

2) What is it really like to develop non-trivial applications with XAML and WPF?

I've experimented with toy examples using Visual Studio 2005 Express plus the extra components needed for .NET 3.0, and viewed tutorials such as this one. Is this an accidental reincarnation of CLIM? It certainly has a familiar feel in some ways, especially the ability to do incremental and interactive development. I'd much prefer a sexpr-based syntax, if not a full-blown Lisp dialect. Does Guile support .NET?

Anyway, I will be devoting a big slice of my free time over the near-term to answer these questions.

5 comments:

Jack Unrue said...

Whoops! Even though Eclipse bug 154116 is somewhat relevant, the Bugzilla ID I actually meant to cite is 154117.

David Lichteblau said...

If you are switching to .NET, I can already imagine Kenny bashing CLIM-GRAPHIC-FORMS as too heavyweight and difficult-to-install before it has even been completed..

:-)

Jack Unrue said...

Yeah, that's potentially a real problem. The .NET runtime is not exactly tiny. A possible counter-argument is that going forward, all of that stuff will be bundled with the OS. A counter-counter-argument could be that many folks have no plans to upgrade beyond XP at least in the near- to mid-term. I could have this argument with myself in fact :-)

I'll keep this issue in mind, thanks!

Anonymous said...

Perhaps you could consider creating a binding to the fox toolkit (fox-toolkit.org),
those Ruby guys are doing it too.

Jack Unrue said...

My goal has been to focus on the native Windows APIs. The extent to which .NET is still kinda-sorta still on User32/GDI is a certainly a consideration as I ponder this potential change.

Nothing against the Fox toolkit per se, but I'm not interested in working on portable GUIs. There are plenty of projects for that (although right at the moment I can't think of a Lisp wrapper for Fox). Otherwise, I would have continued contributing to wxCL.